A controversial clause on telecom ports is at the heart of a debate over the two-year delay of a revised standard used for information technology (IT) equipment. Disagreement over the long-term effects of the delay may mean an uncertain future for the electromagnetic environment in Europe.

After much debate and some heavy industry lobbying, CENELEC has shifted the date of withdrawal (DOW) of EN 55022:1998, which references CISPR 22 (1997), to August 2003. The standard, "Information Technology Equipment—Radio Disturbance Characteristics—Limits and Methods of Measurement," has been under scrutiny worldwide since the discovery that no currently available impedance stabilization network (ISN) conforms to the specifications in clause 9.5 of the third edition of CISPR 22 (1997). ISNs measure the conducted common-mode emissions at telecom ports.

"CENELEC has delayed the date by two years in order to allow for the corrections, currently under way, to be included in the standard," says Mark Bogers of the European Commission (EC). "The main problem with the standard seemed to be ambiguities in the description of the test procedures."

The delay in the withdrawal will likely mean that IT manufacturers could continue marketing older systems such as Ethernet without redesigning the modem cards.

"Hewlett-Packard and all IT manufacturers supported the shift due to the uncertainty of the obtained results," says Rainer Gehrmann, manager of technical regulations and EMC standardization for Hewlett-Packard GmbH. "Our desire was to avoid legal uncertainty, retesting of products, and the likelihood of fixes to existing products."

"Ultimately this [standard] may lead to unnecessary and unjustified product removals from the market. This will generate additional unnecessary cost for industry, authorities, and the consumer," AeA Europe says in a recent position paper. AeA is a high-tech trade association formerly known as the American Electronics Association.

A delay of any length, however, has some experts concerned that such a short-term business advantage will come at the expense of polluting the radio spectrum with excessive emissions.

"A two-year delay in implementation merely puts back the panic by two years," says Tim Williams, who wrote a guidance document addressing the problem for the EMC Test Labs Association (EMCTLA). "Once a solution is identified and documented, I would expect it could be implemented by manufacturers, including product recall and modification, in about three months."

In particular, the delay could adversely affect telecommunications technologies such as digital subscriber lines. Some in the industry are concerned that without this test in place, the huge rollout of xDSL planned in Europe will take place without any legal method for testing its emissions, and many think that the effect will be that the electromagnetic environment in Europe will be permanently ruined. It has been suggested that without effective emissions controls in place at the EU level, some member states may choose to implement their own national restrictions on xDSL and similar technologies. Under the EMC Directive, it may be possible for member states to opt for this route.

"The most significant outcome of a delay in implementation of these limits relates to protection of the EM environment. This clause in EN 55022 is the only product-related limit available for controlling the pollution caused by the rapid rollout of broadband data services such as ISDN and xDSL," says Williams, an EMC consultant with Elmac Services. He says that the only other action that can be taken to preserve the spectrum availability is to enforce system-level radiated emissions limits. In April 2000, for example, the Radiocommunications Agency in the United Kingdom issued MPT 1570, "Radiation Limits and Measurement Standard," to address electromagnetic radiation for telecommunication systems such as DSL that operate in the frequency range of 9 kHz to 300 MHz.

"The EC [members] are naturally not in favor of this, but by accepting a delay, they are shooting themselves in the foot and damaging the credibility of the EMC Directive," Williams asserts.

Others disagree. "It doesn't affect the rollout of xDSL in Europe, which, by the way, is progressing rapidly," Bogers of the EC counters. "Manufacturers should, of course, independently of the standard, ensure that xDSL products don't cause undue interference when installed. I expect that they [xDSL manufacturers] will meet the protection levels contained in the new standard anyway."

Delaying the transition date of the entire standard may be an unnecessary solution to the problems found in a single clause. EMCTLA suggests that the delay "throws away the whole revision because of a bad clause." Other industry groups agree. Among the criticisms of the delay is that the alternative methods presented for testing suspect equipment without using ISNs are acceptable and apply to only a few cable types, and that with appropriate information, test labs should be able to work with the existing standard.

According to the EMCTLA guidance document, the standard is open to various interpretations in terms of test methods, which could result in excessive variation and lack of repeatability between test labs. Considerable controversy surrounds the suitability of currently available ISNs as specified in the third edition of CISPR 22. It is unclear which of two methods of calibration of longitudinal conversion loss (LCL) best represents the properties of actual cable installations.

"The alternative methods are not as technically sound as using properly characterized ISNs, and it has been shown many times that on cables in which you can also use an ISN, the alternative methods do not provide equivalent results," Williams explains. "However, they are better than no test at all, and they are necessary when you have a cable that cannot be tested with an ISN. Most of Technical Guidance Note 42 is about guidance on how to apply these methods in areas where the original standard is confusing."

In May, AeA Europe expressed its concerns to CENELEC about the questionable LCL specifications and calibration methods. "Authorities agree that the situation is not at all acceptable for the test labs and their customers," AeA says. AeA's supporting documentation states that problems caused by premature publication of the standard will be identified at the 2001 IEEE Symposium on EMC.

The document, known as "HP Comments on CISPR 22:1997/EN 55022:1998 clause 9.5," notes that the telecom port requirement has been controversial since publication of the first committee draft because test equipment and calibration methods were not available from commercial sources at that time. It was not until two years after the publication of CISPR 22:1997 that ISNs became available, according to the document. However, the HP document also recommended postponing only implementation of clause 9.5, not the whole standard, until the problems are resolved.

At an expert meeting in Kolberg, Germany, in November 2000, it was also pointed out that no ISN is available that fully conforms to CISPR 22. The German agency RegTP was asked to propose to the European Commission that it postpone the DOW of EN 55022. According to the minutes of that meeting, "Measurement uncertainties of the order of >6 dB make the procedure so inaccurate that it is no longer actually worth making the measurements." Compliance with a tolerance of ±3 dB for the 80 dB LCL specification requires capacitances (to produce the defined LCL) of the order of < 0.1 pF, "which is beyond the limits of feasibility, in particular in respect to serial manufacturing."

By the end of January 2001, two CISPR/G drafts had been withdrawn until the calibration issue could be resolved between CISPR/A and CISPR/G. According to a report from CISPR/G, it was noted that test results likely depend on the manufacturer of the ISN and the method used to calibrate it. The report called this "clearly an unacceptable situation."

The EMCTLA guidance document encourages test laboratories to base their interpretation of the standard and choice of test method on the information provided in the document. "It is as far as possible consistent with the existing standard and with the interpretations known to be under active discussion in CISPR/G," says the document. The document includes a table with tests ranked in order of preference depending on the type of cable. A preliminary flowchart already circulated by CISPR/G is appended to the guidance.

For addressing EN 55022/CISPR 22, ISN manufacturer Schaffner relies on a modified version of the flowchart developed by CISPR/G to determine which product and method to use in each circumstance, according to Ernst Rüegg of Schaffner EMV AG (Luterbach, Switzerland). It is unclear what effect the delay of the standard will have on ISN manufacturers. According to the EMCTLA guidance, however, one ISN manufacturer has already recalled its ISNs for modification.

"The main problems with EN 55022, which were up for discussion at CISPR/G (now reformed as CISPR I) as CDV206 and CDV207, were not resolved at a CISPR meeting held in Bristol [UK] in June," says Dave Imeson, secretary of EMCTLA and director of Compliance Europe Ltd. "There will be new CDVs [committee draft votes]. Hence, it may take all of the two-year delay to resolve the standard."

"There are undoubtedly technical problems, but CISPR/G is well aware of them and is working on solutions, which could be fast-tracked in Europe, if necessary," Williams says.

In a document dated June 2001, however, CENELEC had proposed that the withdrawal date of the entire standard be delayed by two years. "It was also noted that the appropriate solution at present under preparation at CISPR level would take two years before endorsement at CENELEC level and subsequent listing in the Official Journal of the European Community," the document states. It was agreed that the technical board should be asked to shift the DOW of EN 55022:1998 to August 2003.

It is interesting to note that on February 27, 2001, the FCC Office of Engineering and Technology amended sections 15.107 and 15.109 of Chapter 47 of the Code of Federal Regulations to reference CISPR 22:1997. FCC, however, specifically excluded the limits for conducted common-mode disturbance at telecommunication ports. The exclusion was made at the request of the Information Technology Industry Council because "such testing would require the use of stabilization networks that generally are not available." Japan has also adopted the CISPR standard without the controversial clause 9.5 on telecom ports.

Illustration by Taisha Payton

DECT Gets a Boost

Standard EN 301 649 for the digital enhanced cordless telecommunications (DECT) packet radio service has been approved by the ETSI DECT Project and will soon complete the national voting procedure of the European National Standards Organizations. The standard intends to position DECT as a low-cost technology for high-speed radio data transmission in the medium range (50–300 m) for the home environment.

The DECT packet radio service (DPRS) standard defines implementation of packet-data services over the DECT air interface and is the base standard for specifying high-speed data transfer capabilities. Additionally, the document defines requirements on the physical, medium-access control, data-link control, and network layers. Transfer capabilities include high spectrum efficiency due to management entity specification.

"DPRS is used for packet data, which adapts the bit rate to the requested data volume," says Günter Kleindl, chairman of the DECT Project. "DPRS suspends the connection as soon as the data buffer is empty and resumes the connection when new data arrive. DPRS therefore guarantees the efficient use of the spectrum for bursty data traffic."

The first version of DPRS provided four frame-relay services plus one character-oriented service: IEEE 802.3 (Ethernet), IEEE 802.5 (token ring), Internet protocol, and point-to-point protocol, plus V.24 (asynchronous). The new version incorporates high-level modulation options for a maximum operating speed of more than 2 Mb/sec. Additional interworking annexes have also been added to expand the range of supported applications, such as universal serial bus and frame relay Q.922.

To ensure interoperability, TS 101 950, "DPRS Interoperability Test Specification," has also been published, along with application-specific access profiles, each of which identifies a specific application and subset of DPRS services.

In addition to these packet data services, DECT can be used for real-time applications, and further options are on the way. According to Kleindl, the specification for broadband DECT, which would support data transmission from 10 to 20 Mb/sec, should be completed in early 2002.

Illustration by Taisha Payton

Developing the Wireless HUMAN

The Standards Board of the Institute of Electrical and Electronics Engineers (IEEE) Standards Association recently approved the Wireless High-Speed Unlicensed Metropolitan Area Networks (WirelessHUMAN) Project, an initiative within IEEE 802.16 Working Group on Broadband Wireless Access. The initiative is an air-interface standard that will address provision of data services in license-exempt spectra, primarily the 5–6 GHz band.

"Unlicensed spectrum is a huge worldwide market opportunity for fixed broadband wireless access because it may be deployed by any operator without the delay and cost of acquiring a license," said 802.16 chair Roger Marks. "Standardization is key to making this technology readily available to the public as an alternative Internet connection."

To facilitate early standardization of the license-exempt metropolitan area network (MAN) industry, the standard will incorporate modifications of the 802.16 medium-access control layer, and the standard's physical-layer specifications will be based on the orthogonal frequency-division multiplex mechanism of IEEE 802.11a and similar standards.

For more information on Task Group 4, created by 802.16 to draft the standard, visit http://www.WirelessMAN.org or contact Roger Marks at marks@nist.gov.

FCC Streamlines Wireline Equipment Authorization

FCC announced in its Report and Order 00-400, Docket 99-216, the implementation of new procedures for equipment authorization under 47 CFR Part 68. Suppliers of wireline telecommunications equipment may now choose one of two methods to show compliance: approval by a private telecommunications certification body (TCB), which will review test results and issue a grant certifying compliance, or by issuance of a supplier's declaration of conformity (SDoC), as defined by new rules in Section 68.324 and ISO/IEC Guide 22.

The TCB program, begun in June 2000, allows accredited private certification bodies to grant Part 68 certification to telecom equipment. The equipment information and Part 68 certification number then become part of FCC's Part 68 database.

The laboratory used to perform the measurements supporting an SDoC does not require independent accreditation. Instead, an SDoC is submitted to the newly created Administrative Council for Terminal Attachment (ACTA). SDoC holders must retain a description of the testing facilities and procedures used to make the conformance measurements, and the SDoC includes a conformity statement, a list of referenced standards, and the date and location of the declaration. A copy of the SDoC must be made publicly available and must accompany each product manufactured.

ACTA, a private body jointly sponsored by the Telecommunications Industry Association and the Alliance for Telecommunications Industry Solutions, is mandated to work with existing industry standard development organizations (SDOs), in accordance with the consensus process of the American National Standards Institute (ANSI). ACTA will not itself develop standards, but will solicit, review, and approve standards submitted to it by ANSI-accredited SDOs. Excluding requirements related to disabilities and consumer protection issues (hearing-aid compatibility, volume control, fax branding, inside wiring, etc.), ACTA maintains, develops, and publishes the technical rules for most of the current requirements in Subpart D of Part 68.

To download a copy of the report and order, go to http://www.fcc.gov/Bureaus/Common_Carrier/Orders/2000/fcc00400.doc. For further information on Part 68, visit http://www.tiaonline.org/standards/sfg/tr-41.

FYI

An article in CE's 2001 Annual Reference Guide titled "An Overview of ESD Protection Devices" contains material not provided by the author. The information was suggested by reviewers and integrated by CE editors.

The author does not agree with the following references: On page 136 and in Table I, the article states that zener diodes are too slow to protect against ESD. The author states that these diodes are indeed fast enough to react to ESD conditions. The article also states that transient voltage suppression (TVS) diodes are also too slow to protect against nanosecond ESD events. The author states that these are indeed ideal for ESD protection. On page 137, the article states that Schottky and TVS diodes are also too slow to protect against ESD events. The author states that TVS diodes are one of the primary components available in the world for ESD protection.

Back to July/August Table of Contents